Research examining adult, adolescent, and college drinkers suggests that a natural evolution out of problem drinking occurs for a significant portion. Although a fairly substantial body of literature has accrued that examines processes of naturalistic change in adult problem drinkers, this occurrence has rarely been studied among college students. There is converging evidence that the experience (or accumulation) of adverse alcohol events is relevant for this self-change process, but the phenomenological experience of such events has not been studied prospectively. Other important factors related to the progression of problem drinking include peer and parent influences, institutional (college) characteristics, personality characteristics, alcohol expectancies, motivation, and prior experience with alcohol. We propose to conduct a longitudinal study with three waves of data collection (pre-college, freshman, and sophomore year) with two cohorts of students at three colleges in Rhode Island and southern Massachusetts, over sampling for ethnic minority students. In addition to yearly surveys, all students will be regularly screened bi-weekly throughout their freshman and sophomore years using a web-based measure to detect the occurrence of specific alcohol-related adverse events. Students who have events will be randomly selected and then randomly assigned to either an immediate post-event interview or a no-assessment control group. Event interviews will measure students' perception of the event, their attributions, consequences and anticipated consequences, social network response to the event, and intentions to change. Both groups will be interviewed again three months after their events. The design represents an innovative bridging of two strong research methodologies used to investigate college drinking that will allow examination of both protective and risk factors. Resulting knowledge should inform preventive interventions.